Epilepsy is a common neurologic disorder that reflects diverse causes, including genetic factors, brain disease, and generalized or systemic disease. As our knowledge of genetics gradually improves, we now recognize an abundance of findings that show that ion channel gene mutations are intimately associated with epilepsy (1). Among these mutations, those encoding for sodium channels are highly correlated with childhood epilepsy (2). The voltage-gated sodium channel (Nav) is a protein complex consisting of one α subunit and two β subunits and is primarily present in the central nervous system (CNS), peripheral nervous system (PNS), skeletal muscles, and myocardium (3). The α subunit of the Nav channel is a functional subunit of ~2,000 amino acids that is highly conserved. There are nine different Nav α subtypes (Nav1.1–Nav1.9), and mammalian tissues express all nine (4). These subtype proteins are named Nav1.1–Nav1.9 and the corresponding genes are designated SCN1A–SCN5A and SCN8A–SCN11A. Of these, Nav1.1, Nav1.2, Nav1.3, and Nav1.6 are highly expressed in the central nervous system (CNS) and two different subtypes are expressed in muscles (Nav1.4 in adult skeletal muscle and Nav1.5 in myocardium), while Nav1.7, Nav1.8, and Nav1.9 are principally expressed in the peripheral nervous system (5). There are four Nav channel β subunits that are encoded by four genes, i.e., SCN1B–SCN4B (6). Nav channel mutations are related to epilepsy, including α subunits, such as Nav1.1 (SCN1A), Nav1.2 (SCN2A), Nav1.3 (SCN3A), Nav1.6 (SCN8A), and Nav1.7 (SCN9A), and the β subunit (SCN1B) (7). In order to strengthen our understanding of the different sodium channel genotypes and clinical phenotypes, and to guide personalized treatment, determination of disease prognosis, and genetic counseling, we herein summarized the clinical and gene-mutation characteristics of 94 sodium channel gene mutation-related childhood epilepsy patients who were treated at Hunan Children's Hospital. This review, therefore, comprises the broadest analysis of sodium channel genes currently available.

Ion channels constitute the material basis for electrical activity in neurons, and abnormalities in ion channel structure or function can result in electrophysiologic changes that cause abnormal epileptiform discharge. Most of the currently confirmed pathogenic mutations in epilepsy genes encode for ion channels or their auxiliary subunits; of these, gene variants in ion channels comprise the most common cause of genetic epilepsy and sodium channel-related epilepsy and reflect intense areas of research in recent years (11). Sodium channels include nine α subunits and four β subunits, with the SCN1A, SCN2A, SCN3A, SCN8A, SCN9A, SCN11A, and SCN1B encoding human Nav α1, α2, α3, α6, α7, α9, and β1, respectively. Different gene mutations then produce epileptic seizures of varying severity and types.

The SCN1A is located on chromosome 2q 24.3; it contains 31 exons and encodes the Nav1.1 α subunit. As our understanding of the SCN1A improves, we note an increasing number of reports describing SCN1A mutations. The diseases caused by various mutation sites and types are different and range from mild diseases such as GEFS+ and partial epilepsy with febrile seizures plus (PEFS+) to severe diseases such as DS (2). In our study, the 55 SCN1A variant patients encompassed those with relatively mild diseases (four with GEFS+, three with CFS, and two with FS+). With the exception of one patient with a frameshift variant, the remaining patients harbored missense variants. One study showed that the SCN1A mutation was localized to the pedigrees of 5–10% of GEFS+ patients. The GEFS+-related SCN1A mutation is principally concentrated in the intracellular domain of Nav1.1 (12), and almost all mutations are missense mutations, which is consistent with the results from the current study. A previous study indicated that truncation mutations and missense mutations resulted in the most severe clinical symptoms of all SCN1A mutations (13). We herein demonstrated that de novo splicing variants in the SCN1A were present in two patients and in the monozygotic younger sister of one patient, and that these patients died due to status epilepticus. Four patients who failed treatment were DS patients, of whom two harbored missense variants, one had a frameshift variant, and one had a duplication variant.; this, however, was different from previous studies and may have been due to our smaller sample size. A previous study showed that the most common gene mutations in SCN1A-related DS were frameshift mutations, missense mutations, and non-sense mutations; however, splice-site, duplication, and deletion mutations were rarely observed (14). Our study revealed that 40 SCN1A variant patients manifested DS, of whom 16 had missense variants, nine showed frameshift variants, eight had non-sense variants, six possessed splicing variants, and one exhibited a duplication variant; this was generally consistent with previous investigations. Analysis of the treatment responses in 55 patients revealed that CLB, KD, VPA, and TPM produced better efficacy in SCN1A variant-related epilepsy patients, LEV showed slight efficacy, and OXC and LTG were ineffective in controlling convulsions. Symptoms worsened in two of our GEFS+ patients as they experienced recurrent febrile seizures and were ultimately diagnosed with immune-related encephalitis. This suggested that the observed worsening of symptoms during treatment of patients with gene mutation-related epilepsy was not attributable solely to gene mutations, but that other factors such as immune factors and infection should be considered in order to adopt appropriate treatment.

Ever since the first case of SCN2A mutation-induced benign (familial) neonatal/infantile seizure was discovered and confirmed in 2001, the epilepsy spectrum as provoked by SCN2A mutations has continuously expanded to include OS syndrome, infantile spasms, EIMFS, DS, myoclonic-atonic epilepsy, LGS, GEFS+, and other unclassified severe EE (15, 16). Mild epilepsy is primarily related to inherited missense mutations and severe epilepsy is principally precipitated by de novo mutations, with the majority of patients showing delayed mental and motor development (17). In our study, 12 of 14 EE patients harbored de novo variants and manifested moderately severe mental developmental delay. Two children had benign familial epilepsy due to inherited variants, exhibited normal mental and motor development, and responded well to treatment, all of which was consistent with the literature (17). Studies revealed that patients with SCN2A mutation-related EE primarily showed neonatal or infantile onset, and over half showed a neonatal onset (8). In the present study, four EE patients experienced a neonatal onset and two had disease onset before 1–3 months of age. Six patients had a disease onset of 3 months or later, a situation completely different from that described in other reports. Studies showed that the SCN2A mutation generated two Nav1.2 functional changes: gain-of-function (GoF) and loss-of-function (LoF). Early-onset (age at onset < 3 months of age) Nav1.2 is mainly due to GoF and responds well to SCBs. Late-onset (onset ≥ 3 months of age) or non-epilepsy intellectual disability/autism spectrum disorder (ID/ASD) Nav1.2 is mainly due to LoF, responds poorly to SCBs, and may even augur for worsening seizures (8, 18). In our study, we applied sodium channel blockers to four of six patients with early-onset SCN2A-related epilepsy, and OXC and LCS were effective in two patients each. Although sodium channel blockers exhibit some efficacy in early-onset SCN2A-related epilepsy, OXC is not effective in all cases, and LCS and other sodium channel blockers can be used when OXC shows a poor response. We implemented OXC to treat three late-onset SCN2A-related epilepsy patients, and OXC was ineffective in one patient while seizures were completely controlled in the other two. It is obvious, then, that Nav1.2 electrophysiologic changes noted in SCN2A mutation-related EE are more complex. Studies have shown that many patients with epilepsy-related SCN2A mutations present with complex GoF or LoF characteristics, and some rare mutations share both characteristics (19, 20). Hence, it is difficult to apply a simple scheme with respect to classification. Additional experience in future diagnoses and treatments is certainly necessitated to establish more effective methods that can determine GoF or LoF characteristics, so as to produce suitable precision-therapy pharmaceuticals.

SCN8A encodes for the Nav1.6 α subunit and is highly expressed in the central nervous system. Since it was identified as an epilepsy-related gene in 2012, investigators have ascertained that SCN8A mutation-related epileptic seizures are diverse, as they can be manifested as systemic tonic-clonic seizures, spasms, absence, and focal seizures, and can also be present as epileptic syndromes that include BFIE or epileptic encephalopathies, such as Lennox-Gastaut syndrome and West syndrome. In addition to causing epileptic seizures, SCN8A mutations can affect speech, intelligence, and development in pediatric patients (21, 22). In our study, there were nine SCN8A variant-related epilepsy patients, of whom eight showed moderately severe mental and motor impairment. We observed diverse seizure forms but discerned that all patients showed focal seizures. Studies on the functionality of SCN8A mutations revealed that the electrophysiologic changes in the Nav1.6 channel included a persistent increase in Na+ current (INaP), complete channel inactivation, and steady-state rapid inactivation voltage-dependent depolarization transition (23). These three electrophysiologic changes are all GoF alterations of the Nav1.6 channel that increase neuronal excitation due to mutated channels, and correcting the GoF of this channel may therefore constitute a therapeutic target. A related study showed that high-dose sodium channel blockers such as carbamazepine, OXC, and phenytoin inhibited epileptic seizures in patients with the SCN8A mutation (24). In our study, six patients were treated with sodium channel blockers, and convulsions were controlled in one patient after CBZ was added; seizures were controlled in two patients and reduced in one patient; and treatment was ineffective in one patient after OXC treatment. In addition, although seizures decreased in one patient after OXC addition, they still occurred but diminished after LCS was added. This case shows that sodium channel blockers such as CBZ, OXC, and LCS are highly or partially effective in SCN8A variant-related epilepsy. In the present study, the seizure frequency did not decrease or increase in five patients after LEV was applied. VPA was used in seven patients and was effective in three patients and ineffective in four patients. This showed that LEV was ineffective in SCN8A-related epilepsy or that it may have even worsened epilepsy, while VPA exhibited some efficacy consistent with previous studies (25). One study depicted SCN8A-related EE as exhibiting a high mortality rate, reflecting the cause of death as SUDEP primarily. More than 10% of SCN8A mutation patients will develop SUDEP (26, 27), and two of nine patients (22%) died during epileptic seizures in our analysis.

SCN9A is located on chromosome 2q24.3, contains 27 exons, encodes for a 1977- amino-acid Nav1.7 α subunit, and was initially classified as a peripheral nervous system channel gene. It was subsequently reported that Nav1.7 is expressed in the cerebrum, particularly in the fetal hippocampus, showing that it plays a role in the central nervous system (28). Recent reports show that SCN9A mutations are mostly associated with febrile seizures and DS (29). One study showed that SCN9A mutations coexist with SCN1A mutations, that this mutation is a genetic modification factor in SCN1A mutation-related FS+, and that it may worsen the symptoms of mild FS+ (30). As another study showed that the SCN9A variant caused febrile seizures in families without the SCN1A variant (29, 31), it is possible that some SCN9A mutations may portray a modifying role in the presence of more potent mutations, or that it may itself induce mild epileptic seizures (32). There are very few extant reports on epilepsy caused by SCN9A mutations, and the pathogenesis of GEFS+ without an SCN1A mutation remains arcane. In our study, there were six SCN9A variant patients and all manifested fever sensitivity consistent with previous studies. All six patients had unreported missense variants and did not possess SCN1A variants. Of these, four patients were diagnosed with GEFS+, and one patient was diagnosed with CFS; we also reported a patient with non-specific EE. This latter patient began developing recurrent febrile convulsions at 5 months and received VPA as an anticonvulsant treatment. At 7 years and 6 months, seizures were controlled, although the patient exhibited severe developmental delay since infancy. In our study, seizures were controlled after one to two anticonvulsants were applied to six patients, and the response rate with VPA was 100% (six of six). This shows that SCN9A variant-related epilepsy predominantly consists of heterozygous missense variants. Most of the patients reflected fever sensitivity, prominently generalized seizures, and relatively mild symptoms; and anticonvulsants showed good efficacy in these patients.

SCN1B was the first sodium channel gene found to be associated with epilepsy, as Wallace et al. ascertained a point mutation in chromosome 19q13.1 in a GEFS+ pedigree in 1998 (33). Studies showed that heterozygous mutations in this gene were manifested as FS+ and others as non-EE (34), while patients with homozygous recessive mutations presented with DS-like EE symptoms and were at high risk of premature death (35). Compared with classical DS, the clinical symptoms of patients with homozygous SCN1B mutations are more consistent with early infantile developmental and epileptic encephalopathy (DEE), and the latter is more severe than DS (36). We found six SCN1B variant-related epilepsy patients. Three were diagnosed with GEFS+ and three with non-specific EE, and all six patients showed fever sensitivity; the three patients with non-specific EE had moderately mild DS-like symptoms. However, these six patients harbored heterozygous missense variants that were different from those described in previous studies (34, 35). Some investigators demonstrated that there were no differences in surface sodium channel β1 subunit expression in SCN1B mutant (p.Arg 85 Cys) cells compared with wild-type cells. However, the β1 subunit regulates Nav1.1 sodium ion current disappearance, revealing that this SCN1B mutation is a LoF mutation (34), thus providing a basis for the treatment of SCN1B-related epilepsy. With the exception of one patient in our study who did not take anticonvulsants regularly and who exhibited recurrent seizures, seizures were controlled in the other five patients. One patient did not self-administer any anticonvulsant drugs and convulsions did not occur after 6 years of age. Two patients received VPA monotherapy, one patient received diazepam tablets due to convulsion prophylaxis during fever, and another had a poor response to LEV—with seizures decreasing after the patient was switched to VPA and TPM, and controlled after adding LCS. Seizures were ultimately controlled in this patient after the sodium channel blocker LCS was applied, which was different from previous studies that showed that most SCN1B mutations were LoF mutations. Further evaluations of the specific biological mechanisms involved are therefore required.

The SCN3A is located on chromosome 2q24 and encodes for the Nav1.3 channel. Nav1.3 appears to be expressed at different sites at various times. For example, during embryonic and fetal development, Nav1.3 is principally expressed in the central nervous system, with its expression extremely low after birth. During infancy, Nav1.3 is gradually replaced by Nav1.1. SCN3A is also considered to be related to neurodevelopmental disorders and epilepsy, and SCN3A pathogenic mutations reflect many phenotypes. In mild cases, there is a mild cognitive disability, while in severe cases, there is a severe developmental delay due to cerebellar dysfunction (30). We herein noted two SCN3A variant patients, both with cognitive disabilities. The head MRI of these two patients was normal and there were no apparent structural abnormalities. These two SCN3A patients harbored missense variants. However, as our sample size was small, we only had a few SCN3A variant-induced epilepsy cases, and additional in-depth study is thus required.

The SCN11A codes for the sodium channel Nav1.9 is expressed in peripheral pan-sensing neurons and visceral afferent nerves and has been proven to be a threshold channel that is associated with peripheral neuropathy. There is no extant report of SCN11A and epilepsy. One study depicted the SCN11A variant as present in SUDEP patients, but it is unclear whether this variant causes epilepsy or SUDEP (37). We herein found two patients with epilepsy carrying gene variants in SCN11A: one patient was diagnosed with EE, with an unreported c.1646dupA frameshift variant in SCN11A from the patient's mother. The other patient was diagnosed with GEFS+ and had a reported c.2011T>C missense variant in SCN11A from the patient's father, who had a history of febrile seizures when he was young. Nav1.9 is specifically expressed in peripheral nociceptors, and animal models and clinical genetic studies showed that the channel protein plays an important role in inflammatory pain, neurogenic pain, and cold pain but its underlying mechanisms of action in epilepsy remain unelucidated. In our study, we enrolled two cases with SCN11A variants. However, more cases and further genetic functional verification are needed to confirm the relationship between SCN11A and epilepsy.

SCN1A, SCN2A, SCN3A, SCN8A, SCN9A, SCN11A, and SCN1B sodium channel gene mutations were intimately associated with epilepsy and displayed significant heterogeneity in pathogenesis and clinical symptoms. For example, SCN1A mutation-related epilepsy had a disease onset of several months, its clinical presentation was mainly DS, and few patients showed febrile seizures; while sodium channel blockers generally caused symptoms to worsen. SCN2A can cause early-onset encephalopathy within several days after birth or delayed epileptic encephalopathy after 3 months. A majority of seizures exhibited clustering and most sodium channel blockers provided favorable efficacy. Additionally, SCN1A, SCN2A, and SCN8A were associated with SUDEP. The treatment response to antiseizure medications (ASMs) varies depending on the type of Nav channel mutation. Therefore, it is of great significance to study the relationship between genotype, clinical phenotype, and efficacy of ASMs in children with epilepsy. Such studies should provide not only a reference for precise and individualized treatment of epilepsy but also a basis for genetic counseling and prenatal testing to maximize the benefits for children.

The original contributions presented in the study are included in the article/supplementary material, further inquiries can be directed to the corresponding author.

The studies involving humans were approved by the Ethics Committee of Hunan Children's Hospital. The studies were conducted in accordance with the local legislation and institutional requirements. Written informed consent for participation in this study was provided by the participants' legal guardians/next of kin. Written informed consent was obtained from the individual(s), and minor(s)' legal guardian/next of kin, for the publication of any potentially identifiable images or data included in this article.

HF: Data curation, Formal analysis, Funding acquisition, Methodology, Supervision, Writing–original draft, Writing–review & editing. WH: Formal analysis, Writing–review & editing. QK: Investigation, Writing–review & editing. XK: Resources, Writing–review & editing. LWa: Writing–review & editing. XZ: Writing– review & editing. HL: Writing–review & editing. LY: Writing–review & editing. HY: Writing–review & editing. ZJ: Methodology, Writing–review & editing. LWu: Funding acquisition, Project administration, Resources, Writing–review & editing.